APPLICANT'S ABSTRACT: Runaway shelters document high levels of alcohol abuse amongst runaway youth; estimates range from 70% to 85%, with only 15% having ever received treatment for alcohol problems. Research suggests that this population may be unique in the range and intensity of associated problems, including high rates of suicide attempts, pregnancy, prostitution, comorbid diagnoses and criminality. The abuse and neglect experienced by these youth on the streets is compounded by societal neglect in addressing the needs of a population sorely requiring intervention. Shelters for runaways are overcrowded, and many shelters are not equipped to treat youth for alcohol, family and related problems beyond crisis intervention. Most studies to date have collected self-report data on the family and social history. Virtually no research has examined treatment effectiveness in this population. Given the void of treatment outcome research with these youths, and the high level of risk for health and psychological problems, there is a great need for identifying potent interventions. Although research supports the effectiveness of family-based interventions in reducing substance use among adolescents, systematic study of these approaches is sparse, especially when applied to runaway youth, a subpopulation of drinking adolescents. In this proposed work, runaway adolescents will be randomly assigned to: 1) an ecologically-focused, intensive, home-based approach; 2) a traditional office-based approach; or 3) "practice as usual" through the shelter. The efficacy of these approaches in reducing alcohol use, days reducing on the streets, and increasing family and psychological functioning in youth will be examined. Maintenance of treatment effects over time, as well as factors associated with treatment engagement and continuation, will be examined.